Spire debuts AI weather forecasting models built with NVIDIA Omniverse Earth2 tech

Spire debuts AI weather forecasting models built with NVIDIA Omniverse Earth2 tech
by Clarence Oxford
Los Angeles CA (SPX) Mar 19, 2025

Spire Global, Inc. (NYSE: SPIR) has introduced two state-of-the-art artificial intelligence weather models, AI-WX and AI-S2S, which utilize the NVIDIA Omniverse Blueprint for Earth-2 platform. These advanced systems are powered by Spire's exclusive satellite-derived data and offer medium-range and sub-seasonal forecasts designed to support sectors like energy, utilities, and commodities with highly accurate weather predictions.

AI-WX delivers weather forecasts up to 20 days out, while AI-S2S extends forecasting capabilities to 45 days. These models are underpinned by Spire's proprietary data assimilation methodologies, incorporating information from its global satellite network. This includes atmospheric soundings, soil moisture metrics, and ocean wind patterns. Unlike traditional deterministic approaches, Spire's models provide probabilistic forecasts, presenting a spectrum of possible weather scenarios. Clients such as Freepoint Commodities and Dominion Energy are already applying these tools to better assess potential risks and manage the influence of extreme weather on their operations.

"At Freepoint Commodities, staying ahead isn't just about innovation-it's about having an edge in an ever-changing market," said a spokesperson from Freepoint Commodities. "This collaboration with Spire Global represents a major leap forward in weather intelligence for commodity trading. By combining NVIDIA's advanced computing power with Spire's proprietary satellite data, our traders now have the tools to help anticipate market shifts faster, and with increasing accuracy."

As climate change continues to drive more frequent and severe weather events, industries sensitive to weather volatility face growing challenges. Accurate, forward-looking forecasts are critical to mitigating operational disruptions and cost overruns. Spire's AI-driven models equip businesses with actionable insights, allowing them to anticipate weather impacts and respond effectively.

"By harnessing the computational power of NVIDIA GPUs paired with the unique space-based data from our satellite network, we have developed AI-driven models that transform how industries manage weather risks," said Michael Eilts, general manager of weather and climate at Spire. "These models not only predict weather events with unprecedented accuracy, but they also enhance our understanding of weather, enabling energy providers and commodity trading firms to optimize operations and increase both efficiency and profitability in the face of unpredictable weather patterns."

"With Spire's cutting-edge advancements in AI weather modeling, we can anticipate weather impacts further in advance, enabling more precise and strategic risk management for both immediate and long-term weather patterns," said Jeff Mock, lead meteorologist at Dominion Energy. "This heightened foresight is crucial for strengthening resilience against adverse weather, safeguarding critical services, and ensuring operational continuity. By harnessing Spire's advanced forecasting tools and exclusive data sets, we are able to enhance efficiency and minimize disruptions for communities facing the impact of extreme weather."

Earlier, Spire collaborated with NVIDIA to integrate its Radio Occultation data and proprietary data assimilation techniques into NVIDIA Earth-2 Cloud APIs, setting the stage for the development of these new forecasting models.

Research Report:Spire's AI Ensemble Weather Model and Sub-Seasonal AI Model